近日，德国马克斯·普朗克光科学研究所的Clara C.Wanjura与Florian Marquardt合作并取得一项新进展。经过不懈努力，他们利用线性波散射实现全非线性神经形态计算。相关研究成果已于2024年7月9日在国际知名学术期刊《自然—物理学》上发表。

本文提出一种基于线性波散射的神经形态系统的方案，该方案实现了具有高表达性的非线性处理。关键思想是对影响散射过程的物理参数的输入进行编码。此外，研究人员证实训练所需的梯度可以在散射实验中直接测量。研究人员提出了一种基于赛道谐振器的集成光子学实现方案，该方案以最少的波导交叉实现了高连通性。

这项研究工作引入了一种易于实现的神经形态计算方法，可以广泛应用于现有最先进的可扩展平台，如光学，微波和电路。

据悉，用于深度学习应用的神经网络的规模越来越大，它们的能耗也越来越大，因此需要替代神经形态方法，例如使用光学。目前的建议和实现依赖于物理非线性或光电转换来实现所需的非线性激活函数。然而，这些方法在功率水平、控制、能效和延迟方面存在相当大的挑战。

附：英文原文

Title: Fully nonlinear neuromorphic computing with linear wave scattering

Author: Wanjura, Clara C., Marquardt, Florian

Issue&Volume: 2024-07-09

Abstract: The increasing size of neural networks for deep learning applications and their energy consumption create a need for alternative neuromorphic approaches, for example, using optics. Current proposals and implementations rely on physical nonlinearities or optoelectronic conversion to realize the required nonlinear activation function. However, there are considerable challenges with these approaches related to power levels, control, energy efficiency and delays. Here we present a scheme for a neuromorphic system that relies on linear wave scattering and yet achieves nonlinear processing with high expressivity. The key idea is to encode the input in physical parameters that affect the scattering processes. Moreover, we show that gradients needed for training can be directly measured in scattering experiments. We propose an implementation using integrated photonics based on racetrack resonators, which achieves high connectivity with a minimal number of waveguide crossings. Our work introduces an easily implementable approach to neuromorphic computing that can be widely applied in existing state-of-the-art scalable platforms, such as optics, microwave and electrical circuits.

DOI: 10.1038/s41567-024-02534-9

Source: https://www.nature.com/articles/s41567-024-02534-9